Upload
others
View
3
Download
0
Embed Size (px)
Citation preview
Bruggen, Richard Carano and Iqbal S. GrewalPeter Gribling, Jed Ross, Nicolas Van Dhaya Seshasayee, Hua Wang, Wyne P. Lee, inflammatory responseof monocyte effector function and A novel in vivo role for OPGL in activationMechanisms of Signal Transduction:
published online May 15, 2004J. Biol. Chem.
10.1074/jbc.M403968200Access the most updated version of this article at doi:
.JBC Affinity SitesFind articles, minireviews, Reflections and Classics on similar topics on the
Alerts:
When a correction for this article is posted•
When this article is cited•
to choose from all of JBC's e-mail alertsClick here
http://www.jbc.org/content/early/2004/05/15/jbc.M403968200.citation.full.html#ref-list-1
This article cites 0 references, 0 of which can be accessed free at
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
by guest on January 16, 2014
http://ww
w.jbc.org/
Dow
nloaded from
A Novel in vivo Role For OPGL in Activation Of Monocyte Effector Function and Inflammatory Response
Dhaya Seshasayee1, Hua Wang1, Wyne P. Lee1, Peter Gribling1, Jed Ross2,
Nicholas Van Bruggen2, Richard Carano2, Iqbal S. Grewal1
Department of Immunology1 and Department of Physiology2, Genentech, Inc. SouthSan Francisco. CA 94080, USA
RUNNING TITLE: OPGL can activate monocyte effector function
Key words : RANK, cytokines, survival, endotoxic shock, inflammation
Correspondence:Iqbal S. Grewal Ph.D.Department of ImmunologyGenentech Inc.1 DNA Way, MS 34South San Francisco, CA 94080Tel: (650) 225 3239Fax: (650) 225 8221Email: [email protected]
1
JBC Papers in Press. Published on May 15, 2004 as Manuscript M403968200
Copyright 2004 by The American Society for Biochemistry and Molecular Biology, Inc.
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
SUMMARY
Osteoprotegerin Ligand (OPGL) is a member of the TNF ligand superfamily, and has
been shown to be involved in interactions between T cell and dendritic cells. Its role in
monocyte effector function, however, has not been defined. In the present study a role
for OPGL in activating monocytes/macrophages has been characterized. OPGL was
found to up-regulate RANK receptor expression on monocytes, regulate their effector
function by inducing cytokine and chemokine secretion, activate antigen presentation
through up-regulation of co-stimulatory molecule expression, and promote survival.
This activation is mediated through the MAPK pathway as evidenced by activation of
p38 and p42/44 MAPK, and up-regulation of bcl-xL protein levels. A physiological role
for OPGL in monocyte activation and effector function was tested in a model of LPS-
induced endotoxic shock. Administration of RANK-Fc to block OPGL activity in vivo
was able to protect mice from death induced by sepsis, indicating a hitherto un-
described role for OPGL in monocyte function and in mediating inflammatory response.
This was further tested in an animal model of inflammation-mediated arthritis.
Treatment with RANK-Fc significantly ameliorated disease development and
attenuated bone destruction. Thus, our study strongly suggests that administration of
receptor fusion proteins to specifically block OPGL activity in vivo may result in blocking
development of monocyte/macrophage mediated diseases.
2
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
INTRODUCTION
OPGL (Osteoprotegerin ligand) (also called TRANCE (TNF-related activation-induced
cytokine), RANKL (receptor activator of NF-κB ligand), and ODF (osteoclast
differentiation factor)) was discovered almost simultaneously by two groups during
attempts to clone novel genes involved in the regulation of apoptosis and function of
dendritic cells (1,2). OPGL is a member of the TNF ligand superfamily. Most TNF/TNFR
superfamily proteins, including CD40L/CD40, TNF/TNFR, or LT↑β /LT-β receptor, are
expressed in the immune system, and are known to regulate immune response by co-
coordinating homeostasis, T cell activation, dendritic cell function, or the formation of
germinal centers and lymphoid organs such as Peyer’s patches and lymph nodes (3-
5). Sequence analysis has shown that the extracellular domain of OPGL shares 18-
28% amino acid identity with other members of the TNF superfamily and the greatest
identity with CD40L(2). High levels of OPGL mRNA are detectable in T cells in the
lymph nodes and bone osteoblastic cells. OPGL binds to its specific receptor, RANK
(receptor activator of NF-κB), a transmembrane member of the TNFR superfamily (2).
Although RANK mRNA can be detected in skeletal muscle, thymus, liver, colon, small
intestine, and adrenal gland, at the protein level, RANK expression has to date been
detectable only the surfaces of mature dendritic cells and osteoclasts. OPGL has also
been shown to bind an alternate receptor, OPG (osteoprotegerin) (6,7). OPG acts as a
soluble decoy receptor for OPGL and has been shown to neutralize the activity of
3
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
OPGL (8).
Initial functional studies on OPGL revealed its ability to activate dendritic cells
and inhibit apoptosis, resulting in an increase in dendritic cell-mediated T cell
proliferation in an MLR (9). This increase was likely mediated by increased expression
of proinflammatory cytokines such as IL-6 and IL-1, and T cell growth factors such as
IL-12 and IL-15 by dendritic cells (10). In vitro studies also revealed that OPGL, in
combination with CSF-1, could activate mature osteoclasts and mediate
osteoclastogenesis (6,7). Evidence for an essential role for OPGL in the immune
system and bone development was provided by using OPGL-/- mice (11,12). Although
DCs appeared normal, OPGL-deficient mice exhibited defects in early differentiation of
T and B cells, and lacked all lymph nodes. Null mice also exhibited severe
osteopetrosis and a complete lack of osteoclasts. RANK-/- mice expressed similar
phenotypes suggesting that OPGL-RANK interactions provided critical signals
necessary for lymph node organogenesis and osteoclast differentiation (13).
An important connection between bone and the immune system was reported by
Kong et al when they observed that activated T cells could directly activate
osteoclastogenesis through OPGL (14,15). This connection was further strengthened
following a report that this T cell mediated regulation could be suppressed by IFN-γ
produced by T cells themselves (16). These findings suggested that bone metabolism
4
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
is regulated by the immune system through complex and dynamic interactions. In
addition, activation of T cells in vivo could lead to an OPGL-mediated increase in
osteoclastogenesis and bone loss, suggesting that blocking OPGL activity may serve
as an efficient therapeutic approach to attenuate bone loss observed in various
malignant bone disorders (8).
Thus, while a major role for OPGL in regulation of osteoclastogenesis and
chondrocyte differentiation has been established, the exact nature of how OPGL
regulates inflammation and the immune system has not been determined. Osteoclasts
and monocytes are derived from a common myeloid progenitor and are known to utilize
similar signaling pathways involving TRAFs, MAPK and NF-κB, indicating that they
may share a common mechanism of OPGL regulation of their function. This coupled to
the fact that OPGL is most closely related in sequence to CD40L, a molecule crucial to
activation of antigen presenting cells, suggests a role for OPGL in immune response.
In the present study, we have examined a novel role for OPGL in activating
monocytes. RANK protein expression was detected on freshly isolated monocytes, and
treatment with OPGL was shown to activate monocytes, resulting in MAPK activation,
cytokine secretion and up-regulation of co-stimulatory molecule expression. OPGL
was also able to protect monocytes from apoptosis and induced up-regulation of Bcl-2
pro-survival family members such as bcl-xL and bcl-2. These in vitro findings were
confirmed in an in vivo model of LPS-induced septic shock through the use of a
5
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
receptor fusion protein approach to specifically block OPGL activity. Administration of
RANK-Fc was able to protect mice from death induced by sepsis, indicating a novel
role for OPGL in monocyte function in vivo. In addition, treatment with RANK-Fc
significantly ameliorated disease development in an model of inflammation-mediated
arthritis, suggesting therapeutic potential in inflammatory disease. Thus, the
identification of a novel immune cell population regulated by OPGL opens up the
possibility that OPGL may play a key role in inflammatory immune response.
6
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
EXPERIMENTAL PROCEDURES
Cell culture- Monocytes were isolated from human peripheral blood using the
Monocyte Isolation Kit (Milteny Biotec) as per manufacturer’s recommendations.
Briefly, lymphocytes were isolated from human peripheral blood using Lymphocyte
Separation Medium (ICN Pharmaceuticals). A magnetic labeling system (MACS
MicroBeads) was used for the isolation of untouched monocytes from peripheral blood
by depleting non-monocytes. Cells were maintained in complete medium (RPMI 1640
containing 10% heat-inactivated FBS, 50 U/ml penicillin and 50 µg/ml streptomycin)
and cultured at 37oC with 5% CO2.
Reagents- Monoclonal antibodies against p38 MAP kinase, phosphorylated-p38 MAP
kinase, p42/44 MAP kinase, and phosphorylated-p42/44 MAP kinase were purchased
from New England BioLabs Inc. (Beverly, MA). Peroxidase-conjugated α-mouse
secondary antibodies were purchased from Sigma (St. Louis, MO). Monoclonal
antibody to human RANK, recombinant human OPGL and CD40L were purchased from
Alexis Biochemicals (San Diego, CA). PE-α-hCD80, FITC-α-hCD86, PE-α-hClass II,
FITC-α-mouse IgG1antibodies were purchased from BD-Pharmingen (New Jersey,
NJ). LPS was purchased from Sigma (St. Louis, MO). Murine RANK-Fc and control
IgG for in vivo studies were purchased from R&D Systems (Minneapolis, MN) and
7
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
Sigma (St. Louis, MO) respectively.
FACS analyses- Monocytes were adjusted to 5 x 105 cells/ml and incubated in the
presence or absence of OPGL (5 µg/ml) for 24 hours. Cells were then harvested,
washed with PBS containing 2% heat inactivated FBS, incubated with either of the
following antibodies for 15 min at 4oC: PE-conjugated a-hCD80, FITC-conjugated α-
hCD86, PE-conjugated α-hClass II or α-hRANK. Cells stained with α-hRANK were
washed and incubated with FITC-conjugated α-mouse IgG1 antibody for 15 min at
4oC. Following another wash step, cells were analyzed by FACS using CELLQUEST
software (BD BioSciences, San Diego, CA).
Taqman analyses- Monocytes were adjusted to 1 x 106 cells/ml and cultured for 24
hours with (or without) various concentrations of OPGL. Total RNA was then isolated
from OPGL-treated and control cells using TRIzol reagent (Life Technologies,
Gaithersburg, MD). Quantitative RT-PCR analyses were performed with 50 ng of total
RNA sample and 40 µl of a reaction cocktail. The reaction cocktail in the Taqman Core
Kit contained 10x buffer A, 10 Units RNase inhibitor, 200 µM dATP, dCTP, dGTP,
dTTP, 4mM MgCl2, 1.25 Units Taq Gold Polymerase and 25 Units MULV reverse
transcriptase (Perkin Elmer, Foster City, CA). Each well contained a 10µl primer/probe
8
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
mix of 200 nM gene-specific hybridization probe, and 300 nM gene-specific
amplification primers. Thermal cycling conditions: 30 min at 48oC, then 2 min at 50 oC
and 10 min at 95 oC. The reactions then cycled 40 times with 15 sec at 95 oC and 1
min at 60 oC. Reactions and sequence detection was conducted with the ABI Prism
7700 Sequence Detector. Sequence of RANK/GAPDH Taqman primer/probe set used
is as follows:
RANK Forward primer: 5’-AGTGGTGCGATTATAGCCCG-3’
RANK Reverse primer: 5’-GAAGGTTGAGGTGGGAGGATC-3’
RANK Probe: 5’-AGCCTCTAACTCCTGGGCTCAAGCAATC-3’
GAPDH Forward primer: 5’-TGGGCTACACTGAGCACCAG-3’
GAPDH Reverse primer: 5’-CAGCGTCAAAGGTGGAGGAG-3’
GAPDH Probe: 5’-TGGTCTCCTCTGACTTCAACAGCGACAC-3’
Detection of cytokines- Monocytes were induced with various concentrations of OPGL
for 24 hours, and supernatants were harvested. Levels of secreted cytokines in the cell
culture supernatants were determined using ELISAs (BD Pharmingen kits for IL-12 and
IL-6, R & D Systems kits for TNF-α, MIP-1α, IL-1β).
Western blotting- Monocytes (adjusted to 1 x 106 cells/ml) were starved in serum-free
9
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
medium (RPMI1640 containing 50 U/ml penicillin, 50 µg/ml streptomycin) for 4 hours,
and incubated with OPGL for 0, 1, 3,10, 30 and 60 mins. At each time point, cells were
harvested, washed once with PBS, and lysed in buffer (20 mM Hepes, pH 7.4, 2 mM
EGTA, 50 mM -glycerophosphate, 0.1% Triton X-100, 10% glycerol, 1 mM
dithiothreitol, 1 tablet complete protease inhibitor mixture (Roche Molecular
Biochemicals, Indianapolis, IN). Lysates (30 or 50 µg) were separated via SDS-
polyacrylamide gel electrophoresis using 4-20 % Tris-glycine gels (Novex
Electrophoresis) in SDS Running buffer (25 mM Tris, 0.2 M glycine and 3.5 mM SDS),
and transferred onto PVDF membrane (Invitrogen Corp). The membrane was
incubated in blocking buffer followed by primary antibodies for p38 and p42/44 MAPK.
Antibody-antigen complexes were detected using a horseradish peroxidase-
conjugated secondary antibody and ECL system (Amersham Pharmacia Biotech).
Western blotting analyses of bcl-xL and bcl-2 was performed essentially as described
above, with the lysis buffer being 1% SDS, 0.5% Nonidet P-40, 0.15 M NaCl, 10 mM
Tris (pH 7.4), and 1 tablet complete protease inhibitor mixture (Roche Molecular
Biochemicals, Indianapolis, IN).
Survival assays- For in vitro assays, monocytes were adjusted to 5 x 105 c/ml and
incubated with 0.5 mg/ml LPS, 1 µg/ml CD40 ligand, or 1 µg/ml OPGL. At the indicated
10
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
time points, cells in the respective cultures were stained with Annexin V-FITC
(Clontech Labs, Palo Alto, CA) and analyzed by FACS. For in vivo assays, 8 week-old
C57BL/6 mice (The Jackson Laboratory, Bar Harbor, ME) were administered RANK-Fc
or control IgG (100 µg intra-peritonealy) daily starting on day 0, injected with 30 mg
kg-1 body weight of LPS from Escherichia coli serotype O55:B5 (Calbiochem, San
Diego, CA) i.p. on day 1, and monitored for survival.
Induction of antibody-mediated arthritis: Arthritis was induced in two groups of 8-
week-old female Balb/c mice (The Jackson Laboratory, Bar Harbor, ME) by i.v.
injection of 2 mg/mouse (subarthritogenic dose) of a combination of four different
monoclonal antibodies generated by the Arthrogen-CIA® mouse B-hybridoma cell
lines (Chemicon, Temecula, CA). Disease development was aided by an i.p. injection of
50 µg/mouse LPS the following day. Group 1 and 2 were administered RANK-Fc and
control IgG, starting the same day as the injection of the monoclonal antibodies (day 0).
Mice were administered 100 µg of RANK-Fc/control IgG i.p. daily. Mice were sacrificed
on day 14.
X-ray and Micro-CT acquisition- Planar X-ray images of the front and hind paws were
acquired after sacrifice. The paws were severed with an axial cut of the distal
11
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
radius/ulna (front paw) and tibia/fibula (hind paw). The samples were then image with a
digital planar x-ray system (MX-20, Faxitron X-ray, Inc., Wheeling, IL). X-ray images
were acquired with an x-ray tube current of 300 micro-amperes (µA) and a voltage of
26 kilovolts (kV). The extracted mouse samples (front and hind paws) were also
imaged with a µCT40 (SCANCO Medical, Basserdorf, Switzerland) x-ray micro-
computed tomography (µCT) system. A sagittal scout image, comparable with a
conventional planar x-ray, was obtained to define the start and end point for the axial
acquisition of a series of CT images. The location and number of axial images were
chosen to provide complete coverage of the joints of interest for each paw. The target
joints of the front paw were the metacarpophalangeal (MCP) and interphalangeal joints
(PIP and DIP) for digits two through five. The metatarsophalangeal (MTP 2-5), PIP (2-
5), and DIP joints of the hind paw were also evaluated. The µCT images were
generated by operating the x-ray tube at an energy level of 50 kV, a current of 160 µA
and an integration time of 300 milliseconds. Axial images were obtained at an isotropic
resolution of 16 µm.
Radiographic Analysis- X-ray images and µCT were evaluated for bone abnormalities
by application of a semi-quantitative scoring system that was based on the scoring
system for radiographs described by Genant et al (17). A three-dimensional (3D)
surface rendering was created from the µCT data with Analyze (AnalyzeDirect Inc.,
12
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
Lenexa, KS), an image analysis software package. The x-ray image and corresponding
µCT 3D rendering were simultaneously view and evaluated by a single reader (W.L.)
who was blinded to treatment. The degree of erosions and periarticular osteoporosis
were graded with a 6-point score ranging from 0 (normal) to 5 (severe). A cumulative
score was determined for each paw by summing the scores of the individual joints: front
paw (MCP, PIP and DIP, digits 2-5) and hind paw (MTP, PIP, DIP, digits 2-5).
Histological Analysis- Joints from all 4 feet from each animal were scored in three
categories (synovial, bone, and cartilage changes) and added to achieve a final total
score. Scores were derived as follows:
1 = severity is minimal and distribution is multifocal-to-diffuse OR severity is mild but
distribution is focal
2 = severity mild/distribution diffuse OR severity moderate but distribution is focal
3 = severity is moderate and distribution is multifocal
4 = severity is moderate and distribution is diffuse OR severity is severe but distribution
is focal
5 = severity is severe and distribution is multifocal-to-diffuse
Criteria for lesion severity scores:
Synovium, bone and cartilage lesions were scored separately. Synovial lesions were
based on the amount of synovial proliferation (pannus) and inflammation. Bone lesions
13
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
were based on the amount of bone destruction/loss (occasional evidence of
osteoclastic activity in areas of pannus or inflammation as minimal to segmental
complete loss of bone and replacement by pannus or new/noncortical bone as severe.
Cartilage lesions were based on the amount of cartilage destruction/loss (loss of nuclei
and preservation of smooth cartilage surface as minimal to complete fragmentation
and/or loss as severe).
14
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
RESULTS
RANK is expressed on monocytes and is up-regulated upon stimulation with OPGL-
To investigate potential roles for OPGL in monocyte and macrophage effector function,
we first wanted to determine if RANK is expressed on monocytes. Towards this end,
monocytes were freshly isolated from peripheral blood, stained with a monoclonal
antibody to RANK and analyzed by FACS. As shown in Figure 1A, RANK expression
on the cell surface of resting monocytes was clearly detected over control isotype
staining. To determine if OPGL can regulate expression of RANK, monocytes were
treated with various concentrations of OPGL (0, 0.3, 0.6, 1.25 and 2.5 µg/ml) for 24
hours. Total RNA was isolated and quantitative RT-PCR analyses were performed
(Figure 1B). Upon stimulation with OPGL, a dose-dependent increase in RANK mRNA
expression was observed. The highest concentration of OPGL used, 2.5 µg/ml,
stimulated a four-fold increase in RANK expression over un-stimulated monocytes
after 24 hours. This up-regulation in RANK expression was specific to OPGL and was
not observed when monocytes were stimulated with LPS or CD40L (data not shown).
Consistent with an increase in mRNA levels, up-regulation of RANK protein cell
surface expression was also observed in FACS analyses upon OPGL treatment after
24 hours (Figure 1C).
OPGL induces effector function of monocytes by up-regulating secretion of cytokines
15
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
and chemokines- Activation of monocytes by physiological stimuli such as LPS and
TNF family members like CD40L has been shown to induce their effector function
resulting in secretion of pro-inflammatory cytokines such as IL-1β and TNF-α. To test
if OPGL could functionally activate monocytes, we stimulated monocytes for 24 hours
with incremental doses of OPGL and looked for production of pro-inflammatory
cytokines such as TNF-α and IL-1β, T cell-activation cytokines including IL-12 and
IL-6, and chemokines such as MIP-1α. As shown in Figure 2A-E, OPGL was able to
induce secretion of these cytokines from freshly isolated monocytes in a dose-
dependent manner. At the highest concentration of OPGL used (5 µg/ml), levels of
cytokine concentration in the supernatant were 213 pg/ml for IL-12, 7704 pg/ml for IL-
6, 13.4 pg/ml for TNF-α, 803 pg/ml for IL-1β and 8740 pg/ml for MIP-1α. These levels
were comparable to those induced by LPS at 5 µg/ml (373 pg/ml for IL-12, 6193 pg/ml
for IL-6, 21.3 pg/ml for TNF-α, 658 pg/ml for IL-1β and 10395 pg/ml for MIP-1α).
OPGL activates antigen presentation function of monocytes by inducing expression of
co-stimulatory molecules- In addition to playing a dominant role in innate immunity by
phagocytosis of microorganisms, monocytes and macrophages play a very important
role in activating the adaptive immune response by activating T lymphocytes. This is
accomplished by the secretion of activation cytokines such as IL-12 and IL-6, by
presenting antigens, and by providing co-stimulatory signals to T cells. Antigen
16
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
presentation is accomplished by molecules such as MHC Class II, while members of
the B7 family such as B7.1/CD80 and B7.2/CD86 are involved in T cell co-stimulation.
To study if OPGL is involved in activation of above functions, freshly isolated
monocytes were incubated with OPGL (5 µg/ml) and analyzed by FACS. As shown in
Figure 3, OPGL was able to efficiently up-regulate expression of CD80 (A), CD86 (B)
and MHC Class II (C) molecules on monocytes after a 24-hour stimulation. OPGL was
also able to moderately up-regulate CD40 expression on monocytes (data not shown).
The above results thus indicate a novel role for OPGL in activating various in vivo
functions of monocytes and macrophages such as pro-inflammatory cytokine secretion
and T cell activation.
OPGL activates the MAPK pathway in monocytes- To determine if monocyte activation
by OPGL involves the MAPK pathway, we induced monocytes with OPGL (1 µg/ml) for
1, 3, 10, 30 and 60 mins and looked for activation of MAPK pathways (Figure 4).
Phosphorylation of p38 MAPK (A) and p42/44 ERK (B) was observed at 1 min. and
peaked at 10 mins for p38, while it remained at elevated levels for p42/44 MAPK even
at 60 min. Interestingly, phosphorylation of p52/54 JNK/SAPK could not be detected in
monocytes upon OPGL treatment (data not shown) suggesting a differential utilization
of MAPK pathways.
17
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
OPGL enhances survival of monocytes- Based on previous studies suggesting a role
for OPGL as a survival factor for dendritic cells, we induced apoptosis in monocytes by
serum withdrawal and looked for protection from apoptosis induced by OPGL. Annexin
V staining analyses revealed that OPGL protected monocytes from apoptosis at levels
comparable to LPS (0.5 mg/ml) and CD40L (1 µg/ml) (Figure 5A). Similar results were
also obtained for monocytes treated with α-Fas antibody and dexamethasone (data not
shown). To determine possible mechanisms of OPGL-induced survival in monocytes,
we looked for induction of anti-apoptotic proteins belonging to the bcl-2 family. Robust
activation of bcl-xl protein expression was observed starting at 4 hours post-OPGL
stimulation (1 µg/ml) in monocytes, and a moderate but reproducible up-regulation of
bcl-2 protein expression was also observed (Figure 5B).
Blocking OPGL activity results in decreased susceptibility to endotoxic shock- The
above results suggest an important role for OPGL in the activation of effector function
of monocytes during inflammation and this was tested in an in vivo model of LPS-
induced endotoxic shock (18). A soluble receptor form, RANK-Fc, comprising the
extracellular domain of murine RANK was used to block OPGL function in vivo. RANK-
Fc was chosen based on its ability to bind OPGL exclusively, as opposed to the soluble
decoy receptor OPG, which can also bind an alternate TNF ligand, TRAIL (19). 8 week
old C57/Bl6 mice were injected intra-peritonealy with 30 mg/kg LPS and 100 µg
18
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
RANK-Fc or control IgG and monitored for survival. As shown in Figure 6,
administration of RANK-Fc was able to significantly block death induced by septic
shock. After 72 hours, while only 20% of mice in the control group survived, 80% of
RANK-Fc-treated mice were still alive. These results indicate a dominant intrinsic role
for OPGL in mediating inflammatory response and potential clinical applications in
combating septic shock.
Therapeutic potential for blocking OPGL activity in inflammatory arthritis- Based
on the above-observed ability of RANK-Fc to block OPGL function in activating
monocytes/macrophages, we tested roles for RANK-Fc in blocking development of
inflammation-mediated arthritis in vivo (20,21). This model is distinct from conventional
collagen-induced arthritis (CIA) models in that inflammation induced by LPS is one of
the major determinants of disease development. In this model, 8 week-old Balb/c mice
were injected with a cocktail of four different monoclonal antibodies to collagen on day
0, and 50 µg of LPS on day 1 to activate inflammation. Mice were treated with 100 µg
RANK-Fc or control IgG daily and monitored for inflammation in the joints and arthritis.
As shown in Figure 7A, swelling in the joints of mice administered RANK-Fc was
significantly lower than that in control IgG-treated mice, which was also reflected in the
histology scores (Figure 7B). Mice were sacrificed on day 14 representing the end of
the study, and radiographic examinations were performed. Digital planar x-rays of
19
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
fore- and hind paws revealed that the affected joints in control IgG–treated mice had
severe osteolysis and osteophyte production accompanied by disfigurement and this
was absent in the RANK-Fc–treated mice (Figures 7C, D). Disease was scored using
conventional radiographic methods (scale of 0 to 4, with 4 representing most severe
disease) with x-rays and is shown in Table 1. Individual paw scores for RANK-Fc
treated mice range from 0 to 3, while those for Ig-treated mice are primarily 4 or 3. In
addition, joints were also scanned by micro-CT to generate 3-dimensional rendered
images. X-rays and micro-CT renderings were simultaneously viewed and visually
evaluated for bone destruction by the application of a modified radiographic scoring
system (17). Development of arthritis was significantly inhibited in RANK-Fc treated
mice as evidenced by suppressed bone erosion and loss of bone density (Figure 8 A).
The degree of erosions and periarticular osteoporosis were graded with a 6-point score
ranging from 0 (normal) to 5 (severe). A cumulative score was determined for each paw
by summing the scores of the individual joints: front paw (MCP, PIP and DIP, digits 2-
5) and hind paw (MTP, PIP, DIP, digits 2-5), and is graphed for each mouse (five per
group) in control IgG and RANK-Fc treated groups (Figure 8B). Mice treated with
RANK-Fc had significantly reduced bone destruction (mean score 6 + 2.5) compared
to the control group (mean score 33.4 + 11.3). These results thus demonstrate good
efficacy for RANK-Fc in treatment of inflammation-mediated arthritis.
20
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
DISCUSSION
Interactions between the TNF superfamily ligands and their cognate receptors are
essential in regulating immune response, both by promoting cell survival and
proliferation and through activation-induced cell death (3). OPGL and its receptor
RANK, members of the TNF superfamily, are known to be key regulators of bone
metabolism and essential for the development and function of osteoclasts (6,7).
Osteoclasts and monocytes are derived from a common myeloid progenitor indicating
that there may be common mechanisms of regulation of their function (22). In addition,
OPGL is very closely related to CD40L (1,2), which is known to be crucial for activation
of monocytes/macrophages (23). This suggests a potential role for OPGL in regulation
of effector function by these cells. In the present study, we have defined a novel role for
OPGL in activating monocytes/macrophages. OPGL was found to regulate their
effector function by inducing cytokine and chemokine secretion, activate antigen
presentation through up-regulation of co-stimulatory molecule expression, and inhibit
serum withdrawal-induced apoptosis. These in vitro findings were confirmed in an in
vivo model of LPS-induced septic shock through the use of a receptor fusion protein
approach to block OPGL activity. Administration of RANK-Fc was able to protect mice
from death induced by sepsis, indicating a novel important role for OPGL in monocyte
function and in mediating inflammatory response. This was further tested in an animal
21
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
model of antibody-induced arthritis. Treatment with RANK-Fc significantly ameliorated
disease development and attenuated bone destruction, suggesting therapeutic potential
in inflammatory disease.
OPGL and its receptor RANK were initially discovered based on their expression
in T cells and dendritic cells respectively (2,9). Interactions between ligand and receptor
were shown to be important for activation of dendritic cell function and in regulating
cross talk between T cells and dendritic cells. Dendritic cells serve primarily to present
antigens to T cells and form an important part of the antigen presenting cell population.
Monocytes/macrophages represent another important component of the APC
population, and in addition, also perform key effector functions such as inducing
inflammation and directly killing microorganisms. They also serve as an important
bridge between innate and adaptive immune responses by priming T cells through
secretion of activation cytokines and presentation of antigens. To date, OPGL has not
been implicated in the regulation of this key immune population and roles for OPGL in
activation of monocyte/macrophages function have not been defined. Our results
describe a novel function for OPGL in monocyte activation and define important roles in
regulating their diverse functions including secretion of pro-inflammatory and T cell
activation cytokines, antigen presentation and co-stimulation.
Results from our studies indicate that the biochemical pathways involved in
OPGL signaling in monocytes appear to be mediated through p38 MAPK and p42/44
22
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
ERK. In mammalian cells, at least three MAPK pathways have been identified: ERK1/2,
JNK1/2/SAPK and p38 MAPK pathways. Various cell growth and differentiation stimuli
have been shown to activate the ERK1/2 pathway, leading to proliferation and
differentiation responses (24). The JNK1/2 and p38 MAPK pathways are primarily
activated by inflammatory cytokines and environmental stress and lead to inflammatory,
apoptotic or developmental responses (24). Previous studies in monocytes have linked
T cell-dependent secretion of the pro-inflammatory cytokines, IL-1β and TNF-α, to
activation of the ERK1/2 pathway (25). Our studies suggest that this T cell-dependent
signal may be mediated through OPGL, and the subsequent ERK1/2 activation in
monocytes may play a central role in the secretion of pro-inflammatory cytokines.
Circulating monocytes have a limited life span and when recruited to a site of
inflammation, will undergo apoptosis in the absence of further survival stimuli. CD40L,
present on CD4+ T cells has been shown to inhibit apoptosis of circulating monocytes
and promote their survival (26). Our results indicate that OPGL may also be important
in vivo for monocyte survival. A key parameter that determines whether a cell will
respond to an apoptotic signal is the ratio of death antagonist (bcl-2, bcl-xL, bcl-w,
mcl-1, bfl-1) to agonists (bax, bak, bcl-xS, bad, bid, bim) belonging to the Bcl-2 family
(27). In this context, up-regulation of bcl-xL and bcl-2 induced by OPGL may well
account for the decrease in the number of monocytes undergoing apoptosis at the site
23
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
of inflammation in vivo. Enhanced survival of inflammatory cells, including monocytes,
may be an important factor in the establishment of chronic inflammation that
characterize both atopic and autoimmune diseases (28,29). Thus OPGL may play a
role in the persistence of inflammatory responses associated with these disorders by
prolonging monocyte/macrophage survival.
Classical activation signals for monocytes/macrophages include LPS, and
another TNF member, CD40L. LPS is a component of cell walls of gram-negative
bacteria and is a potent stimulator of inflammatory response. CD40, the receptor for
CD40L is expressed on primarily on B cells, antigen presenting cells and endothelial
cells (30-32). In monocytes/macrophages, activation of CD40 by its ligand, CD154, has
been shown to induce secretion of pro-inflammatory cytokines and chemokines such
as IL-1β, IL-6, IL-8, IL-10, IL-12, TNF-α and MIP-1α (33,34). Activation of CD40
signaling in monocytes/macrophages also results in up-regulation of co-stimulatory
molecules such as B7.1/B7.2 for T cell activation (34), nitric oxide generation (35) and
induction of metalloproteinase production (36) for killing of microorganisms. Thus, there
are many similarities in the various physiological roles for CD40L and OPGL, including
expression patterns of ligand/receptors and effects on monocytes/macrophages.
Results from our and previous studies strongly indicate that they may play important
roles in mediating the primary immune response, through direct killing and the induction
of inflammation. However, OPGL has been implicated as an immediate early gene in T
24
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
cells (1) and may be acting prior to CD40L during innate and adaptive immune
response. Rapid up-regulation of OPGL upon activation of the T cell receptor on T cells
could specifically activate antigen-presenting cells and promote their survival. Both
antigen-specific T cells and the antigen presenting cells would therefore depend on
each other for activation and survival. Mature APCs that fail to present antigen to T
cells would not receive T cell help and would therefore die of neglect. This feedback
loop may play an important in the initiation and maintenance of immune response.
Previous studies have used the decoy receptor OPG as a therapeutic reagent
for various malignant bone disorders (37-39). OPG however, is also known to bind
another TNF ligand, TRAIL, thus complicating interpretation of results. In our study, we
have used RANK-Fc to block OPGL activity since RANK binds to OPGL exclusively.
Results from both in vivo models of septic shock and inflammation-mediated arthritis
described in our study demonstrate the therapeutic efficacy of RANK-Fc in these
applications, establishing a dominant role for OPGL in mediating sepsis and
inflammatory arthritis. Furthermore, the ability of RANK-Fc to block
monocyte/macrophage function strongly indicates that it may be ameliorate disease in
additional clinical indications where these cell populations are known to play major
roles.
AcknowledgementsWe wish to thank Dan Tumas for help with histology analysis.
25
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
REFERENCES1. Wong BR, R. J., Arron J, Robinson E, Orlinick J, Chao M, Kalachikov S, Cayani
E, Bartlett FS 3rd, Frankel WN, Lee SY, Choi Y. (1997) J Biol Chem 272, 25190-25194
2. Anderson DM, M. E., Billingsley WL, Dougall WC, Tometsko ME, Roux ER, Teepe MC, DuBose RF, Cosman D, Galibert L. (1997) Nature 390, 175-179
3. Locksley, R. M., Killeen, N., and Lenardo, M. J. (2001) Cell 104, 487-501
4. Fu, Y. X., and Chaplin, D. D. (1999) Annu Rev Immunol 17, 399-433
5. Grewal, I. S., and Flavell, R. A. (1998) Annu Rev Immunol 16, 111-135
6. Lacey DL, T. E., Tan HL, Kelley MJ, Dunstan CR, Burgess T, Elliott R, Colombero A, Elliott G, Scully S, Hsu H, Sullivan J, Hawkins N, Davy E, Capparelli C, Eli A, Qian YX, Kaufman S, Sarosi I, Shalhoub V, Senaldi G, Guo J, Delaney J, Boyle WJ. (1998) Cell 93, 165-176
7. Yasuda H, S. N., Nakagawa N, Yamaguchi K, Kinosaki M, Mochizuki S, Tomoyasu A, Yano K, Goto M, Murakami A, Tsuda E, Morinaga T, Higashio K, Udagawa N, Takahashi N, Suda T. (1998) Proc Natl Acad Sci U S A 95, 3597-3602
8. Hofbauer LC, K. S., Dunstan CR, Lacey DL, Boyle WJ, Riggs BL. (2000) J Bone Mineral Res 15, 2-12
9. Wong BR, J. R., Lee SY, Sauter B, Li HL, Steinman RM, Choi Y. (1997) J Exp Med 186, 2075-2080
10. Josien R, W. B., Li HL, Steinman RM, Choi Y. (1999) J Immunol 162, 2562-2568
11. Kong YY, Y. H., Sarosi I, Tan HL, Timms E, Capparelli C, Morony S, Oliveira-dos-Santos AJ, Van G, Itie A, Khoo W, Wakeham A, Dunstan CR, Lacey DL, Mak TW, Boyle WJ, Penninger JM. (1999) Nature 397, 315-323
12. Li J, S. I., Yan XQ, Morony S, Capparelli C, Tan HL, McCabe S, Elliott R, Scully S, Van G, Kaufman S, Juan SC, Sun Y, Tarpley J, Martin L, Christensen K,
26
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
McCabe J, Kostenuik P, Hsu H, Fletcher F, Dunstan CR, Lacey DL, Boyle WJ. (2000) Proc Natl Acad Sci U S A 97, 1566-1571
13. Dougall WC, G. M., Charrier K, Rohrbach K, Brasel K, De Smedt T, Daro E, Smith J, Tometsko ME, Maliszewski CR, Armstrong A, Shen V, Bain S, Cosman D, Anderson D, Morrissey PJ, Peschon JJ, Schuh J. (1999) Genes Dev 13, 2412-2424
14. Kong YY, F. U., Sarosi I, Bolon B, Tafuri A, Morony S, Capparelli C, Li J, Elliott R, McCabe S, Wong T, Campagnuolo G, Moran E, Bogoch ER, Van G, Nguyen LT, Ohashi PS, Lacey DL, Fish E, Boyle WJ, Penninger JM. (1999) Nature 402, 304-309
15. Kong YY, B. W., Penninger JM. (1999) Immunol Cell Biol 77, 188-193
16. Takayanagi H, O. K., Hida S, Chiba T, Murata S, Sato K, Takaoka A, Yokochi T, Oda H, Tanaka K, Nakamura K, Taniguchi T. (2000) Nature 408, 600-605
17. Genant, H. K., Jiang, Y., Peterfy, C., Lu, Y., Redei, J., and Countryman, P. J. (1998) Arthritis Rheum 41, 1583-1590
18. Ashkenazi, A., Marsters, S. A., Capon, D. J., Chamow, S. M., Figari, I. S., Pennica, D., Goeddel, D. V., Palladino, M. A., and Smith, D. H. (1991) Proc Natl Acad Sci U S A 88, 10535-10539
19. Emery, J. G., McDonnell, P., Burke, M. B., Deen, K. C., Lyn, S., Silverman, C., Dul, E., Appelbaum, E. R., Eichman, C., DiPrinzio, R., Dodds, R. A., James, I. E., Rosenberg, M., Lee, J. C., and Young, P. R. (1998) J Biol Chem 273, 14363-14367
20. Terato, K., Harper, D. S., Griffiths, M. M., Hasty, D. L., Ye, X. J., Cremer, M. A., and Seyer, J. M. (1995) Autoimmunity 22, 137-147
21. Terato, K., Hasty, K. A., Reife, R. A., Cremer, M. A., Kang, A. H., and Stuart, J. M. (1992) J Immunol 148, 2103-2108
22. Ross, F. P. (2000) J Clin Invest 105, 555-558
23. Stout, R. D., Suttles, J., Xu, J., Grewal, I. S., and Flavell, R. A. (1996) J Immunol 156, 8-11
27
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
24. Schaeffer, H. J., and Weber, M. J. (1999) Mol Cell Biol 19, 2435-2444
25. Suttles, J., Milhorn, D. M., Miller, R. W., Poe, J. C., Wahl, L. M., and Stout, R. D. (1999) J Biol Chem 274, 5835-5842
26. Suttles, J., Evans, M., Miller, R. W., Poe, J. C., Stout, R. D., and Wahl, L. M. (1996) J Leukoc Biol 60, 651-657
27. Kroemer, G. (1997) Nat Med 3, 614-620
28. Vignola, A. M., Chanez, P., Chiappara, G., Siena, L., Merendino, A., Reina, C., Gagliardo, R., Profita, M., Bousquet, J., and Bonsignore, G. (1999) J Allergy Clin Immunol 103, 563-573
29. Ravirajan, C. T., Pittoni, V., and Isenberg, D. A. (1999) Int Rev Immunol 18, 563-589
30. Stamenkovic, I., Clark, E. A., and Seed, B. (1989) Embo J 8, 1403-1410
31. Clark, L. B., Foy, T. M., and Noelle, R. J. (1996) Adv Immunol 63, 43-78
32. Van Kooten, C., and Banchereau, J. (1996) Adv Immunol 61, 1-77
33. Stout, R. D., and Suttles, J. (1996) Immunol Today 17, 487-492
34. Kiener, P. A., Moran-Davis, P., Rankin, B. M., Wahl, A. F., Aruffo, A., and Hollenbaugh, D. (1995) J Immunol 155, 4917-4925
35. Tian, L., Noelle, R. J., and Lawrence, D. A. (1995) Eur J Immunol 25, 306-309
36. Malik, N., Greenfield, B. W., Wahl, A. F., and Kiener, P. A. (1996) J Immunol 156, 3952-3960
37. Akatsu, T., Murakami, T., Ono, K., Nishikawa, M., Tsuda, E., Mochizuki, S. I., Fujise, N., Higashio, K., Motoyoshi, K., Yamamoto, M., and Nagata, N. (1998) Bone 23, 495-498
38. Capparelli, C., Kostenuik, P. J., Morony, S., Starnes, C., Weimann, B., Van, G., Scully, S., Qi, M., Lacey, D. L., and Dunstan, C. R. (2000) Cancer Res 60, 783-787
28
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
39. Pearse RN, S. E., Yaccoby S, Wong BR, Liau DF, Colman N, Michaeli J, Epstein J, Choi Y. (2001) Proc Natl Acad Sci U S A 98, 11581-11586
29
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
TABLE 1
RANK-Fc Treatment ameliorates development of arthritis.
# of Paws with arthritic score 0 1 2 3 4
Control IgG 0 0 0 9 15
RANK-Fc 4 6 5 9 0
Distribution of radiographic scores: Paws from control IgG and RANK-Fc treated mice were scored on a scale 0 to
4, with 4 representing most severe bone destruction.
30
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
FIGURE LEGENDS
FIG 1. RANK is expressed on monocytes and is upregulated upon OPGL treatment . A,
Freshly isolated monocytes were stained for RANK expression (bold line). Isotype
control staining is indicated in grey. B, Monocytes were treated with (or without)
indicated concentrations of OPGL for 24 hours, and total RNA was isolated.
Quantitative RT-PCR analyses were performed, and fold-increase in RANK mRNA
levels of OPGL-treated over control untreated cells is shown. GAPDH levels were used
to normalize loading. C, Monocytes were analyzed for RANK protein expression at 0
hrs and 24 hours after incubation in the presence or absence of OPGL (5 µg/ml). RANK
cell surface expression at 0 and 24 hours is shown in gray and bold lines respectively.
FIG 2. OPGL activates effector function of monocytes by inducing cytokine and
chemokine secretion. Freshly isolated monocytes were treated with the indicated
concentrations of OPGL for 24 hours and supernatants were harvested. Levels of (A)
IL-12, (B) IL-6, (C) TNF-α, (D) IL-1β and (E) MIP-1α in the supernatants were
determined by ELISA.
FIG 3. OPGL upregulates expression of co-stimulatory molecules on monocytes .
Freshly isolated monocytes were analyzed for (A) CD80, (B) CD86 and (C) MHC Class
II protein expression by FACS at 0 hours and 24 hours after incubation in the presence
31
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
or absence of OPGL (5 µg/ml). Co-stimulatory molecule expression at the cell surface
at 0 and 24 hours is shown in gray and bold lines respectively.
FIG 4. OPGL activates p38 MAPK and p42/44 ERK pathways in monocytes . Freshly
isolated monocytes were serum-starved for 6 hours, and treated with OPGL (1 µg/ml).
At the indicated time points, cells lysates were harvested and Western blotting analyses
performed for detection of phosphorylated (A) p38 MAPK and (B) p42/44 MAPK
pathways.
FIG 5. OPGL protects monocytes from apoptosis and induces expression of pro-
survival Bcl-2 proteins. A, Freshly isolated monocytes were incubated in serum-free
medium in the presence of OPGL (1 µg/ml), CD40L (1 µg/ml), LPS (0.5 mg/ml) or no
stimulus control. At the indicated time points, cells were harvested and apoptotic cells
were detected by Annexin V- FITC staining. Cell stimulated with OPGL are indicated
by filled squares, with CD40L and LPS by circles and triangles respectively, while
control unstimulated cells are represented by open squares. B, Freshly isolated
monocytes were treated with OPGL (1 µg/ml) for the indicated time points. Cells were
lysed and western blotting analyses performed for detection of bcl-xL and bcl-2 protein
levels. Levels of actin at each time point were assayed to control for loading.
32
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
FIG 6. Blocking OPGL activity with RANK-Fc protects mice from LPS-induced
endotoxic shock. 8 week-old C57/Bl6 mice were injected i.p. with 30 mg/kg LPS
serotype O55:B1 and 100 µg/day RANK-Fc (or control IgG) and monitored for survival.
Group 1 administered RANK-Fc is represented by black squares and group 2 (control
IgG) by white squares.
FIG 7. RANK-Fc inhibits inflammation and blocks disease development in an
inflammation-mediated model of collagen-induced arthritis. 8 week-old Balb/c mice
were injected with a combination of LPS (50 µg/mouse) and a cocktail of four α-
collagen antibodies (4 mg/mouse). Mice were administered 100 µg/day RANK-Fc or
control IgG and disease progression was monitored daily. A, LPS-mediated
Inflammation per paw was scored (on a scale of 0 to 4, 4 being most severe) daily
based on joint swelling. The average score per paw is plotted as a function of days-
post α-collagen antibody injection. B, Inflammation was also measured by histological
scoring of individual paws (on a scale of 0 to 3 representing none, mild, moderate and
severe states of inflammation). C, A radiograph from a representative control mouse
showing signs of arthritis, which is significantly reduced in the RANK-Fc-treated
mouse. Arrowheads indicate sites of osteolysis. A blowup of this region is shown in (D).
33
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
FIG 8. RANK-Fc blocks arthritis by inhibiting bone erosion and loss of bone density in
an inflammation-mediated model of arthritis. Mice in control IgG- and RANK-Fc -
treated groups were sacrificed 14 days after α-collagen antibody injection and joints
were scanned by micro-CT. A, Bone erosion and loss of bone density in joints of mice
representative of RANK-Fc and control IgG groups are shown. The images are a 3D-
surface rendering created from the µCT data using Analyze image analysis software. B,
X-ray images and µCT were evaluated for bone abnormalities by application of a
semi-quantitative scoring system that was based on the scoring system for radiographs
described by Genant et al. The x-ray image and corresponding µCT 3D rendering were
simultaneously view and evaluated by a single reader who was blinded to treatment.
The degree of erosions and periarticular osteoporosis were graded with a 6-point score
ranging from 0 (normal) to 5 (severe). A cumulative score was determined for each paw
by summing the scores of the individual joints: front paw (MCP, PIP and DIP, digits 2-
5) and hind paw (MTP, PIP, DIP, digits 2-5).
34
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from
by guest on January 16, 2014http://w
ww
.jbc.org/D
ownloaded from